Coronavirus disease 2019 (COVID-19) is an infectious disease caused by a new type of coronavirus: severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The outbreak first started in Wuhan, China in December 2019. The first kown case of COVID-19 in the U.S. was confirmed on January 20, 2020, in a 35-year-old man who teturned to Washington State on January 15 after traveling to Wuhan. Starting around the end of Feburary, evidence emerge for community spread in the US.
We, as all of us, are indebted to the heros who fight COVID-19 across the whole world in different ways. For this data exploration, I am grateful to many data science groups who have collected detailed COVID-19 outbreak data, including the number of tests, confirmed cases, and deaths, across countries/regions, states/provnices (administrative division level 1, or admin1), and counties (admin2). Specifically, I used the data from these three resources:
JHU (https://coronavirus.jhu.edu/)
The Center for Systems Science and Engineering (CSSE) at John Hopkins University.
World-wide counts of coronavirus cases, deaths, and recovered ones.
NY Times (https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html)
The New York Times
``cumulative counts of coronavirus cases in the United States, at the state and county level, over time’’
COVID Trackng (https://covidtracking.com/)
COVID Tracking Project
``collects information from 50 US states, the District of Columbia, and 5 other US territories to provide the most comprehensive testing data’’
Assume you have cloned the JHU Github repository on your local machine at ``../COVID-19’’.
The time series provide counts (e.g., confirmed cases, deaths) starting from Jan 22nd, 2020 for 253 locations. Currently there is no data of individual US state in these time series data files.
Here is the list of 10 records with the largest number of cases or deaths on the most recent date.
Next, I check for each country/region, what is the number of new cases/deaths? This data is important to understand what is the trend under different situations, e.g., population density, social distance policies etc. Here I checked the top 10 countries/regions with the highest number of deaths.
The raw data from Hopkins are in the format of daily reports with one file per day. More recent files (since March 22nd) inlcude information from individual states of US or individual counties, as shown in the following figure. So I turn to NY Times data for informatoin of individual states or counties.
The data from NY Times are saved in two text files, one for state level information and the other one for county level information.
The currente date is
## [1] "2020-04-09"
First check the 10 states with the largest number of deaths.
## date state fips cases deaths
## 2084 2020-04-09 New York 36 159937 7067
## 2082 2020-04-09 New Jersey 34 51027 1700
## 2074 2020-04-09 Michigan 26 21375 1076
## 2070 2020-04-09 Louisiana 22 18283 702
## 2055 2020-04-09 California 6 20191 548
## 2065 2020-04-09 Illinois 17 16422 534
## 2073 2020-04-09 Massachusetts 25 18941 503
## 2102 2020-04-09 Washington 53 9608 456
## 2061 2020-04-09 Georgia 13 10885 412
## 2057 2020-04-09 Connecticut 9 9784 380
For these 10 states, I check the number of new cases and the number of new deaths. Part of the reason for such checking is to identify whether there is any similarity on such patterns. For example, could you use the pattern seen from Italy to predict what happen in an individual state, and what are the similarities and differences across states.
Next I check the relation between the cumulative number of cases and deaths for these 10 states, starting on March
First check the 10 counties with the largest number of deaths.
## date county state fips cases deaths
## 44818 2020-04-09 New York City New York NA 87028 5150
## 44817 2020-04-09 Nassau New York 36059 20140 778
## 44424 2020-04-09 Wayne Michigan 26163 10093 504
## 44845 2020-04-09 Westchester New York 36119 17004 389
## 44837 2020-04-09 Suffolk New York 36103 17413 369
## 43842 2020-04-09 Cook Illinois 17031 11415 351
## 44745 2020-04-09 Bergen New Jersey 34003 8343 345
## 44750 2020-04-09 Essex New Jersey 34013 6069 312
## 45734 2020-04-09 King Washington 53033 3888 260
## 44405 2020-04-09 Oakland Michigan 26125 4247 246
For these 10 counties, I check the number of new cases and the number of new deaths.
The positive rates of testing can be an indicator on how much the COVID-19 has spread. However, they are more noisy data since the negative testing resutls are often not reported and the tests are almost surely taken on a non-representative random sample of the population. The COVID traking project proides a grade per state: ``If you are calculating positive rates, it should only be with states that have an A grade. And be careful going back in time because almost all the states have changed their level of reporting at different times.’’ (https://covidtracking.com/about-tracker/). The data are also availalbe for both counties and states, here I only look at state level data.
Since the daily postive rate can fluctuate a lot, here I only illustrae the cumulative positave rate across time, for four states with grade A data. Of course since this is an R markdown file, you can modify the source code and check for other states.
## R version 3.6.2 (2019-12-12)
## Platform: x86_64-apple-darwin15.6.0 (64-bit)
## Running under: macOS Catalina 10.15.4
##
## Matrix products: default
## BLAS: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRblas.0.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRlapack.dylib
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] httr_1.4.1 ggpubr_0.2.5 magrittr_1.5 ggplot2_3.2.1
##
## loaded via a namespace (and not attached):
## [1] Rcpp_1.0.3 pillar_1.4.3 compiler_3.6.2 tools_3.6.2
## [5] digest_0.6.23 evaluate_0.14 lifecycle_0.1.0 tibble_2.1.3
## [9] gtable_0.3.0 pkgconfig_2.0.3 rlang_0.4.4 yaml_2.2.1
## [13] xfun_0.12 gridExtra_2.3 withr_2.1.2 dplyr_0.8.4
## [17] stringr_1.4.0 knitr_1.28 grid_3.6.2 tidyselect_1.0.0
## [21] cowplot_1.0.0 glue_1.3.1 R6_2.4.1 rmarkdown_2.1
## [25] purrr_0.3.3 farver_2.0.3 scales_1.1.0 htmltools_0.4.0
## [29] assertthat_0.2.1 colorspace_1.4-1 ggsignif_0.6.0 labeling_0.3
## [33] stringi_1.4.5 lazyeval_0.2.2 munsell_0.5.0 crayon_1.3.4